Dominic Boutet

Dendrites endow artificial neural networks with accurate, robust and parameter-efficient learning - Nature Communications Artificial neural networks, central to deep learning, are powerful but energy-consuming and prone to overfitting. The authors propose a network design inspired by biological dendrites, which offe...

Dendrites doing AI problems, so great to see the increased activity! www.nature.com/articles/s41...

It can run basic processing steps like band pass filtering, time freq stuff like hilbert/fft/wavelet/PAC, connectivity stuff like AEC/PLI/PLV, etc. It is packaged and available through pypi using "python -m pip install pystorm3". 3/3

It does not handle interactions with Brainstorm or any database and mainly operates as a library of functions to perform operations on data (sensor signals and imaging kernel). It handles parcellation; just pass in the imaging kernel as a list of matrices grouped by parcel. 2/3

GitHub - NeuroLife77/pystorm: Python implementation of various Brainstorm functions, often written for GPU or jit speed-up. Python implementation of various Brainstorm functions, often written for GPU or jit speed-up. - NeuroLife77/pystorm

Since my last post sharing tools I've made in the past got some attention, here's another one about a python port for some functions from Brainstorm™ (#notanad) that I made with a focus on runtime and scalability (e.g., numba jit or cuda gpu) if anyone is interested:
github.com/NeuroLife77/.... 1/3

It contains stuff like code for finding nullclines, finding & classifying fixed points (all with associated mathematical derivations), plotting a phase plane. It also has interactive phase plane plots to explore the behavior of the model around some point in the parameter space (e.g., image below).

GitHub - NeuroLife77/Wilson_Cowan_analysis: Analytical derivations and numerical analyses associated with the Wilson-Cowan model Analytical derivations and numerical analyses associated with the Wilson-Cowan model - NeuroLife77/Wilson_Cowan_analysis

I've been working with the Wilson-Cowan neural mass model for a while now and, over time, I've accumulated a bunch of tools to explore/visualize it. I recently decided to collect it all into a github repository available here if anyone is interested: github.com/NeuroLife77/...